In conventional high-accuracy traveling table apparatuses of this type, a workpiece is placed fixedly on a table, which travels in the directions of the X-, Y-, Z-, and .theta.-axes (or sometimes only in the directions of the X- and Y-axes), in order to locate the workpiece accurately. In a traveling table section for locating the workpiece with respect to the X- and Y-directions, as shown in FIGS. 1 and 2, an X-table 14 is provided on a base 10. It can reciprocate linearly along guide ways 12, in the X-direction. A Y-table 18, which is disposed on the X-table 14, can reciprocate along guide ways 16, in the Y-direction, perpendicular to the X-direction. A workpiece fixing member 20, such as a holder, is disposed on the upper surface of the Y-table 18. After a workpiece W is fixed to tables fixing member 20, tables 14 and 18 are caused to travel along the guide ways 12 and 16, in their respective predetermined directions, by drive means (not shown). Thus, the workpiece W on the member 20 can be moved in both the directions of the X-and Y-axes.
An L-shaped laser mirror 22 for X/Y-direction position measurement is mounted on the upper surface of the Y-table 18. Moved distances in the X- and Y-directions are calculated in accordance with the variation of the laser value between the mirror 22 and a laser interferometer (not shown), which is fixed to a baseplate. The X- and Y-tables 14 and 18 are controlled so that the workpiece W is located accurately in accordance with a position measurement, based on the calculated values.
In the prior art apparatuses described above, the workpiece fixing member 20 and the laser mirror 22 for position measurement are mounted directly on the upper surface of the Y-table 18. Thus, regarding the distance between the member 20 and the mirror 22 as invariable, the position of the table 18 is measured by calculating its moved distance, in accordance with the variation of the aforesaid laser value. However, if X- and Y-tables 14 and 18 are deformed elastically or distorted by a temperature rise on the guide-way side when the tables 14 and 18 are moved, the aforesaid distance varies, so that the measured value is subject to errors, correspondingly. This lowers the accuracy of the table-position measurement.
More specifically, in moving the workpiece fixing member 20 in the directions of the X- and Y-axes, fluctuations of pre-load may be caused by variations of the manufacturing accuracy of the guide ways 12 and 16 or of mounting surfaces of the tables 14 and 18 for the guide ways, change of weight balance attributable to the movement of the tables, etc. As a result, a tensile or compressive force is applied to the upper surface of the Y-table 18 on which the fixing chamber 20 and the laser mirror 22 are mounted, so that the mounting surface is extended or contracted. Thus, the distance between the member 20 and the mirror 22 varies, thereby causing errors in the position measurement. In the electron-beam printing apparatuses, in particular, the Y-table 18 is usually formed of a nonmagnetic material, such as aluminum, titanium or beryllium bronze. Since, in particular, aluminum has low rigidity, the Y-table 18 is liable to be deformed while it is traveling. If the variation of the aforesaid distance, caused by the extension or contraction of the surface of the Y-table 18, is about 0.05 .mu.m, it is not insignificant for the high-accuracy traveling table apparatuses of this type, whose positional accuracy at 3 .sigma. is expected to be 0.1 .mu.m or less.
FIG. 3 shows the deformation of Y-table 18 while it is traveling. The degree of extension or contraction of the upper surface of Y-table 18, caused by such deformation, can be given by EQU E=4t.delta./l, (1)
where E is the maximum degree of extension or contraction of the surface of the Y-table 18, t is the table thickness, l is the table length, and .delta. is the maximum distortion of the table surface, with respect to the vertical direction or Z-direction. If t=30 mm, l=250 mm, and .delta.=0.2 .mu.m, then E=4t.delta./l=0.1 .mu.m. Thus, the surface of the Y-table 18 is extended by 0.1 .mu.m at the maximum. If the phase of the distortion is inverted vertically, then the table surface is contracted by 0.1 .mu.m. Such extension or contraction of the surface of the Y-table 18 directly influences the variation of the distance between the object fixing member 20 and the laser mirror 22.
If the surface of the Y-table 18, carrying the laser mirror 22, is extended or contracted, the flatness of the mirror 22, which usually ranges from about .lambda./20 to .lambda./10 (.lambda. is a wavelength of the laser beam; .lambda..apprxeq.0.6 .mu.m for Ne-He laser), is undesirably deformed. Thus, the measurement of the moved position is subject to errors, so that the accuracy of the apparatus is lowered.
In a conventional arrangement, the slit 24 is tentatively formed in a part of the Y-table 18 so that its opposite side edges are located above and below, as shown in FIG. 1. The slit 24 serves to absorb the working errors of the guide ways 16 and the X- and Y-tables 14 and 18. This arrangement is based on the so-called pre-load system, utilizing elasticity. In this case, however, the region around the slit 24 it is low in rigidity, and is susceptible to vibration.